Objective: Venous thromboembolic event (VTE) after spine surgery is a rare but potentially devastating complication. With the advent of machine learning, an opportunity exists for more accurate prediction of such events to aid in prevention and treatment. Methods: Seven models were screened using 108 database variables and 62 preoperative variables. These models included deep neural network (DNN), DNN with synthetic minority oversampling technique (SMOTE), logistic regression, ridge regression, lasso regression, simple linear regression, and gradient boosting classifier. Relevant metrics were compared between each model. The top four models were selected based on area under the receiver operator curve; these models included DNN with SMOTE, linear regression, lasso regression, and ridge regression. Separate random sampling of each model was performed 1000 additional independent times using a randomly generated training/testing distribution. Variable weights and magnitudes were analyzed after sampling. Results: Using all patient-related variables, DNN using SMOTE was the top-performing model in predicting postoperative VTE after spinal surgery (area under the curve [AUC] =0.904), followed by lasso regression (AUC = 0.894), ridge regression (AUC = 0.873), and linear regression (AUC = 0.864). When analyzing a subset of only preoperative variables, the top-performing models were lasso regression (AUC = 0.865) and DNN with SMOTE (AUC = 0.864), both of which outperform any currently published models. Main model contributions relied heavily on variables associated with history of thromboembolic events, length of surgical/anesthetic time, and use of postoperative chemoprophylaxis. Conclusions: The current study provides promise toward machine learning methods geared toward predicting postoperative complications after spine surgery. Further study is needed in order to best quantify and model real-world risk for such events.

OBJECTIVE: Isolated, nondisplaced skull fractures (ISFs) are a common result of pediatric head trauma. They rarely require surgical intervention; however, many patients with these injuries are still admitted to the hospital for observation. This retrospective study investigates predictors of vomiting and ondansetron use following pediatric ISFs and the role that these factors play in the need for admission and emergency department (ED) revisits.METHODS: The authors identified pediatric patients (< 18 years old) with a linear ISF who had presented to the ED of a single tertiary care center between 2008 and 2018. Patients with intracranial hemorrhage, significant fracture displacement, or other traumatic injuries were excluded. Outcomes included vomiting, ondansetron use, admission, and revisit following ED discharge. Both univariable and multivariable analyses were used to determine significant predictors of each outcome (p < 0.05).RESULTS: Overall, 518 patients were included in this study. The median patient age was 9.98 months, and a majority of the patients (59%) were male. The most common fracture locations were parietal (n = 293 [57%]) and occipital (n = 144 [28%]). Among the entire patient cohort, 124 patients (24%) had documented vomiting, and 64 of these patients (52%) received ondansetron. In a multivariable analysis, one of the most significant predictors of vomiting was occipital fracture location (OR 4.05, p < 0.001). In turn, and as expected, both vomiting (OR 14.42, p < 0.001) and occipital fracture location (OR 2.66, p = 0.017) were associated with increased rates of ondansetron use. A total of 229 patients (44%) were admitted to the hospital, with vomiting as the most common indication for admission (n = 59 [26%]). Moreover, 4.1% of the patients had ED revisits following initial discharge, and the most common reason was vomiting (11/21 [52%]). However, in the multivariable analysis, ondansetron use at initial presentation (and not vomiting) was the sole predictor of revisit following initial ED discharge (OR 5.05, p = 0.009).CONCLUSIONS: In this study, older patients and those with occipital fractures were more likely to present with vomiting and to be treated with ondansetron. Additionally, ondansetron use at initial presentation was found to be a significant predictor of revisits following ED discharge. Ondansetron could be masking recurrent vomiting in ED patients, and this should be considered when deciding which patients to observe further or discharge.

Introduction: Distal medium vessel occlusion (DVO) thrombectomy has been shown to be efficacious with safety profiles comparable to large vessel occlusion (LVO) thrombectomy. A novel, highly-trackable, bevel-tipped Zoom 35 catheter can be used as an aspiration catheter for DVO thrombectomy. Methods: This is a retrospective, single-arm, multi-institutional observational study evaluating the efficacy and safety of aspiration thrombectomy for DVO using the Zoom 35 catheter. Patient demographics, presenting and discharge NIHSS, primary and rescue thrombectomy, site of occlusion, TICI score, and intracranial hemorrhage were chart abstracted. Descriptive statistics were used to evaluate the efficacy and safety of thrombectomy. Results: Fourteen patients (mean age 66.64 ± 13.75 years) were included. The mean NIHSS at presentation was 10.79 ± 5.48, and the mean ASPECTS was 9.00 ± 0.89. Nine patients (64.3%) received tPA. Primary occlusion location was M3 in nine cases (64.3%), M2/M3 junction in two cases (14.3%), A2 in one case (7.1%), A3 in one case (7.1%), and P1 in one case (7.1%). TICI scores were 3 in seven cases (50.0%), 2C in three cases (21.4%), and 2B in four cases (28.6%). There was one postoperative SAH (7.1%) and one asymptomatic ICH (7.1%). Mean discharge NIHSS was 3.38 ± 4.44, with a mean decrease of 7.31 from presentation (p < 0.0001, t-test). Conclusion: Zoom 35 beveled-tip aspiration microcatheters are highly trackable and associated with improved radiographic and clinical outcomes for the treatment of DVO with a good safety profile.